Stress inoculation therapy

Stress Inoculation Therapy (SIT) (Meichenbaum Reference Meichenbaum1985) involves a form of cognitive behavioural therapy, and is based on the theory that behaviour and feelings are determined by individual perceptions, and can therefore be cognitively reframed. SIT is intended to break the ‘stress cycle’ of negative reactions and ineffective responses to stressful events by teaching a range of coping skills and behaviours. The three-stage process involves patients reconceptualising their responses to specific stressors, before learning coping skills and applying these to imaginary (role-played) and real-life stressors (Meichenbaum & Cameron Reference Meichenbaum, Cameron, Meichenbaum and Jaremko1983).

SIT was used by a treatment in 27 of the papers reviewed, and several randomised controlled trials identified beneficial effects (Law et al. Reference Law, Logan and Baron1994; Szabo & Marian Reference Szabo and Marian2012; Varker & Devilly Reference Varker and Devilly2012; Hourani et al. Reference Hourani, Tueller, Kizakevich, Lewis, Strange, Weimer, Bryant, Bishop, Hubal and Spira2016; Navaee & Kaykha Reference Navaee and Kaykha2019), with some positive effects still present two years after the intervention (Hourani et al. Reference Hourani, Tueller, Kizakevich, Lewis, Strange, Weimer, Bryant, Bishop, Hubal and Spira2016). However, in the context of orangutan rehabilitation, the cognitive reframing element of SIT would be extremely difficult to achieve in a non-linguistic form and may be counterproductive in blunting responses to stressors, for example by promoting the inhibition of appropriate reactions to stressful events during the ‘skills acquisition’ phase of the treatment (Meichenbaum & Cameron Reference Meichenbaum, Cameron, Meichenbaum and Jaremko1983). It is important for caretakers to be able to identify stressors for orangutans by reading their behaviour, to avoid further harm. This is already a challenge, as non-human primates are adept at masking overt expressions of pain and illness (Pelsker & Mayer Reference Pelsker and Mayer2008). SIT has also received criticism for moving the locus of control away from the patient (towards the therapist), and for framing certain responses to stress as ‘irrational’ (Hurley et al. Reference Hurley, Barrett and Reet2006). As will be discussed later in Controllability and predictability of stress, a sense of agency and control over the environment is thought to have a beneficial effect for a range of taxonomic groups. Although SIT is not appropriate here, certain coping skills involved in the treatment may be beneficial, for example problem-solving skills.

Inoculation from major lifetime stressors

Alongside daily stressors related to life in captivity, all orangutans in rehabilitation will have experienced at least one major life stressor, i.e. capture and travelling to the rehabilitation centre. However, depending on the prior history of each orangutan, they are likely to have experienced additional major stressors leading up to their rescue, such as witnessing the death of their mother, as well as stressors during their time in rehabilitation, for example veterinary interventions. Although some studies identified a positive linear relationship between cumulative lifetime adversity and the odds of being diagnosed with a stress-related disorder (Gerber et al. Reference Gerber, Frankfurt, Contractor, Oudshoorn, Dranger and Brown2018; Fernandez et al. Reference Fernandez, Choi, Marshall, Vicente, Saldivia, Kohn, Koenen, Arheart and Buka2020), there is more robust evidence for a U-shaped relationship. In a large study involving a stratified random-digit-dial telephone survey with 2,398 members of the public, Seery et al. (Reference Seery, Holman and Silver2010) found that individuals with some experience of lifetime adversity had lower ratings for distress, functional impairment, and PTSD symptoms, higher ratings for life satisfaction, and were least affected by recent adversity than those with no experience of prior adversity, or those with high exposure to adversity (categorised as mean + 1 SD; ̴ 87th percentile). These results are supported by Seery et al. (Reference Seery, Leo, Lupien, Kondrak and Almonte2013), who reported a similar U-shaped relationship between lifetime adversity and pain-induced catastrophising and negative affect, with moderate numbers of prior major life stressors (two to seven) associated with the highest resilience to stress. If a U-shaped relationship between life stressors and resilience exists for orangutans, it is very important that an ‘upper limit’ of life stressors is established, so it can be avoided. However, because orangutans’ life histories prior to arrival at the rehabilitation centres are often unknown, this will be difficult to achieve. The concept of stress inoculation and the effects of prior adversity will be discussed further in the Animal intervention section of this paper.

Controllability and predictability of stress

Uncontrollable stress can lead to ‘learned helplessness’ (see S1 Appendix; Supplementary material), a term which describes the failure to respond to avoidable shocks, due to prior experience of inescapable traumatic shock (Overmier & Seligman Reference Overmier and Seligman1967). This should be taken seriously in the context of orangutan rehabilitation, where space is limited and orangutans have very little control or ability to make decisions to change their situation. For example, they are restricted in what and when they eat, who they have social contact with, and where they can go while in the rehabilitation centre. Maier (Reference Maier2001) found that inescapable stress followed by repeated exposure to the specific context in which stress took place prolonged depression and learned helplessness in rats indefinitely. Captive orangutans may be repeatedly exposed to stressful contexts such as relatively small enclosures or visits from the vet. Repeatedly ‘refreshing’ the stress associated with captivity and small environments may have long-lasting detrimental effects. It would be beneficial to identify coping strategies currently used by orangutans in rehabilitation to manage long-term stressors.

There is strong evidence that there are positive effects of being able to control electric shock stress in male rodents, in protecting against stress and depression when exposed to a variety of future stressors and enhancing social exploration and the ability to learn how to discriminate between positively and negatively reinforced odours (see Table 2[a]). Similar benefits are seen in sheep. Greiveldinger et al. (Reference Greiveldinger, Veissier and Boissy2009) found that teaching lambs how to control an airblast to the muzzle during feeding led to lower emotionality and vigilance compared with subjects who did not have the opportunity to control the stressor. Predictability can also reduce the impact of stress. Lambs that experienced a predictable sudden event – either due to regularity of appearance, or because it was associated with a cue – had a reduced startle response and less disruption to feeding than those that experienced an unpredictable event (Greiveldinger et al. Reference Greiveldinger, Veissier and Boissy2007). Although neither predictability nor controllability of the stressor had a noticeable effect on chicken (Gallus domesticus) behaviour, loss of both had the effect of increasing their frustration, measured by higher levels of vocalisations (Zimmerman & Koene Reference Zimmerman and Koene1998). When learned helplessness has occurred, two studies on dogs and rats, respectively, found that it can be reversed by force, i.e. by a human physically dragging the animal away from the source of electric shock (Seligman et al. Reference Seligman, Maier and Geer1968, Reference Seligman, Rosellini and Kozak1975). Although the ethical problems with these methods are substantial, it does indicate that recovery from learned helplessness is theoretically possible.

Direct stress summary

This section clearly demonstrates the potential harms from both acute and chronic forms of stress. However, some forms of milder, less extended, and more controllable forms of stress may have potential benefits. It is natural for orangutans to experience stressful situations in the wild, and it is therefore important for rehabilitant orangutans to be given the opportunity to develop resilience by successfully overcoming a range of different challenges prior to release. Puzzle-boxes, or treatments such as a controllable airblast during feeding, may be worth considering, as they provide a mildly frustrating – but ultimately rewarding – challenge that can be resolved through persistence and problem-solving. The importance of effort-based reward is something that has been addressed in human studies. For example, Dweck (Reference Dweck1975) found a way to ‘treat’ learned helplessness in human children, by teaching them to take responsibility by attributing failures to a lack of effort, rather than ability. Appropriate difficulty levels of cognitive enrichment will be discussed further in Environmental conditions below. There may also be specific personality traits that help to protect against stress, such as low anxiety, low emotionality, and positive affect, which may help orangutans adapt to a new environment. However, although generally low levels of anxiety may be beneficial, it is important that orangutans are still responsive to genuine threats, so that these can be avoided. Therefore, interventions that induce an anxious response to specific threats, for example snakes and humans, can still be utilised, even if overall low anxiety levels are desirable for resilience in orangutans.

Brief separations

Brief daily separation treatments (15–60 min) have a range of beneficial effects, and this is supported by a strong body of evidence. Over the past two decades, a series of stress inoculation studies were carried out on squirrel monkeys (Samiri sciureus) raised in a normal, species-typical way in a laboratory environment. These involved moving subjects from their natal group to an enclosure adjacent to an unfamiliar group for one hour each week, for a period of ten weeks during adolescence, with the purpose of inoculating the monkeys against future stress. This treatment resulted in benefits such as reduced anxiety, increased time spent exploring and interacting with novel objects, and enhanced inhibitory control (Parker et al. Reference Parker, Buckmaster, Schatzberg and Lyons2004, Reference Parker, Buckmaster, Justus, Schatzberg and Lyons2005, Reference Parker, Rainwater, Buckmaster, Schatzberg, Lindley and Lyons2007, Reference Parker, Buckmaster, Lindley, Schatzberg and Lyons2012, Reference Parker, Buckmaster, Hyde, Schatzberg and Lyons2019). However, it should be noted that these studies were all carried out by the same research team, and some research subjects were used for multiple studies, so caution should be taken in generalising the results to other populations. Many rodent studies found positive effects in adulthood of 15 min of daily keeper handling (involving separation from the litter) during infancy or adolescence, including protection against stress, a reduction in anxiety, depression, and fear, and increased playfulness and exploration (see Table 2[b]). These brief separation treatments may be transferable to the context of orangutan rehabilitation. Exposing young infant orangutans to mild separation stress by removing them from their social group and placing them next to unfamiliar orangutans may have stress-inoculating benefits. However, careful management decisions would need to be made on an individual basis regarding the appropriateness of this method. In rehabilitation centres, detailed knowledge of prior traumatic events experienced by individual orangutans is usually not available. As the studies detailed above are carried out on normally reared laboratory animals with known histories, there is less risk of unexpected trauma surfacing during the separations. This potential interaction between prior stress and the effect of separation is related more generally to the nature of the relationship between life stressors and resilience, discussed in the section above on human resilience.

Table 2(b). Summary of maternal/group separation treatment effects by taxonomic group, and sex (96/266 papers reviewed)

(Grey-shaded cells highlight primate studies; S = sex (F = female, M = male, B = both/mixed, ? = not stated)

Longer separations

There is fairly consistent evidence in rodents that longer periods of separation (2–4 h daily) have a range of negative effects (see Table 2[b]), and this is also supported by a primate study (Pryce et al. Reference Pryce, Dettling, Spengler, Spaete and Feldon2004). Wild orangutan infants are in near-continuous bodily contact with their mother for the first two years of life, only gaining full locomotory independence at 5–6, and sharing her nest until weaning at 6–8 years old (van Noordwijk & van Schaik Reference van Noordwijk and van Schaik2005; van Adrichem et al. Reference van Adrichem, Utami, Wich, van Hooff and Sterck2006). Even years after gaining independence, orangutans have been observed to occasionally ‘visit’ their mother at eleven years old (van Noordwijk et al. Reference van Noordwijk, Sauren, Abham, Morrogh-Bernard, Atmoko, van Schaik, Wich, Atmoko, Setia and van Schaik2009). However, early maternal separation is inevitable for nearly all orangutans rescued from the illegal wildlife trade, as the process of capture usually requires the death of the infants’ mother (Nijman Reference Nijman2005). This means that infant orangutans rely mostly on a combination of human caretakers and peer-rearing. A study of laboratory chimpanzees (Pan troglodytes) found that individuals who had been separated from their mothers earlier, and isolated for longer, were less social, less dominant, and more affected by stressful experiences (Reimers et al. Reference Reimers, Schwarzenberger and Preuschoft2007). Compared with maternally raised individuals, same-species surrogacy leads to a reduction in locomotion and time spent exploring in a number of other studies (Sackett Reference Sackett1972; Röder et al. Reference Röder, Timmermans and Vossen1989; Corcoran et al. Reference Corcoran, Pierre, Haddad, Bice, Suomi, Grant, Friedman and Bennett2012). In the context of orangutan rehabilitation, early maternal separation will unavoidably have a detrimental impact. However, social support may help to ameliorate stress; in one marmoset study (Callithrix kuhli), anxiety experienced upon moving to a new enclosure was reduced by the presence of their breeding partner during the transition (Smith et al. Reference Smith, McGreer-Whitworth and French1998).

Prolonged isolation

Orangutans in the illegal wildlife trade may spend prolonged periods of time, even years, with no conspecific contact. However, it may be possible to reverse certain behavioural effects of isolation. Some 1970s primate studies measured the consequences of extreme forms of separation in rhesus macaques. Although disturbing in their methods, two studies have demonstrated successful reversals of the traumatic effects of prolonged isolation. Suomi and Harlow (Reference Suomi and Harlow1972) managed to reverse the effects of six months of isolation in four male rhesus monkeys, by socially exposing them to younger, normally reared female monkeys. After six months of ‘social rehabilitation’, isolated monkeys demonstrated similar social behaviour to controls and an absence of stereotypies. Similar results were achieved in another study of the same species after an entire year of total isolation (Novak & Harlow Reference Novak and Harlow1975). Particularly traumatised orangutans may therefore benefit from peer support in carefully arranged group housing, so that individuals who are coping more effectively can experience a positive effect by association (see Social stress for potential risks). Another technique which has been borrowed from the rehabilitation of ex-laboratory primates is increasing the ratio of caretakers to orangutans in order to monitor progress more carefully, although this may result in excess humanisation (Palmer Reference Palmer2020).

Separation stress summary

There are many potential harms from extended periods of maternal/group separation. However, briefer separations could be used as a mechanism to create a positive, stress-inoculating effect, as seen in the squirrel monkey studies by Parker and colleagues. Maintaining contact with a social peer during separations may help to reduce associated stress. These kinds of ‘gentle’, sporadic separations should be explored as a potentially appropriate treatment for orangutans in rehabilitation, applying a principle of caution and initially trialling with very short periods of time.

Environmental enrichment

Environmental enrichment can have a variety of benefits in rodents, including protecting against the negative effects of future stress on anxiety, depression, fear, and emotional reactivity, as well as increasing activity levels, learning ability, and spatial memory (see Table 2[c]). Cognitive challenges can also have beneficial effects, with effort-based reward training in rats leading to improved persistence and problem-solving (Bardi et al. Reference Bardi, Rhone, Franssen, Hampton, Shea, Hyer, Huber and Lambert2012; Lambert et al. Reference Lambert, Hyer, Rzucidlo, Bergeron, Landis and Bardi2014). This makes sense in terms of having the opportunity to exercise agency (see S1 Appendix; Supplementary material), and overcoming a certain degree of challenge having the effect of eliciting a positive emotional state (Clark Reference Clark2011). There is also strong evidence that environmental enrichment improves the well-being of pigs (Sus scrofa domesticus), including promoting long-term memory, mobility, and diversity of behaviour, and reducing emotionality (De Jong et al. Reference De Jong, Prelle, van De Burgwal, Lambooij, Korte, Blokhuis and Koolhaas2000; Wemelsfelder et al. Reference Wemelsfelder, Haskell, Mendl, Calvert and Lawrence2000; Puppe et al. Reference Puppe, Ernst, Schön and Manteuffel2007). However, the form of enrichment used may be important. In a study comparing the effects of natural and artificial (e.g. plastic/manufactured) novel object enrichment on rats, Lambert et al. (Reference Lambert, Hyer, Bardi, Rzucidlo, Scott, Terhune-Cotter, Hazelgrove, Silva and Kinsley2016) observed that rats interacted nearly three times more with natural enrichment items than artificial enrichment items matched according to their functionality and purpose (i.e. climbing, shelter, or manipulation), as well as having a greater reduction in anxiety-like behaviours. Although natural enrichment may be preferable, the practicalities of this with larger animals can be complicated, as enrichment and enclosure features must also be durable and not easily destroyed. Chappell and Thorpe (Reference Chappell and Thorpe2022) argue that non-natural enclosure modifications can simulate the mechanics of a natural environment and provide similar physical and cognitive challenges to those provided by wild environments.

Table 2(c). Summary of environmental conditions treatment effects by taxonomic group, and sex (41/266 papers reviewed)

(Grey-shaded cells highlight primate studies; S = sex (F = female, M = male, B = both/mixed, ? = not stated)

However, moving from an enriched to a barren environment can be more harmful than only experiencing barren environments (Bolhuis et al. Reference Bolhuis, Schouten, Schrama and Wiegant2006). In studies of pigs, moving from an enriched to a barren environment has been found to reduce activity levels (Bolhuis et al. Reference Bolhuis, Schouten, Schrama and Wiegant2006), and increase pessimism in a cognitive bias test used to measure attitudes towards an ambiguous cue (Douglas et al. Reference Douglas, Bateson, Walsh, Bédué and Edwards2012) Similarly, starlings (Sturnus vulgaris) who had moved from enriched to barren environments had increased pessimism in a cognitive bias test (Bateson & Matheson Reference Bateson and Matheson2007). This has implications for orangutan rehabilitation, as orangutans are likely to experience a fluctuation of environmental conditions, from wild environments to extremely poor captive conditions (for illegally traded individuals), to improved conditions at the rehabilitation centre with varying levels of enrichment, depending on the stage of rehabilitation. It is not clear from the literature how a series of multiple changes in condition would affect individuals, and this makes it very difficult to be sure how changing environmental conditions are likely to impact orangutans in rehabilitation. However, it seems likely that an increasing trajectory of environmental complexity would be beneficial, with orangutans in rehabilitation experiencing gradual increases in enrichment and autonomy and, as far as possible, avoiding a reversal in conditions.

Environmental stress/deprivation

Most evidence suggests that environmental deprivation, for example wire floors and restricted bedding during infancy, has a range of negative effects in rodents, for example increased anxiety, impaired spatial memory and novel object recognition, and reduced exploration (see Table 2[c]). However, the studies on environmental stress in birds and fish had less straightforward results. In zebra finches (Taeniopygia guttata), a low quality diet led to faster spatial learning and exploration compared to individuals on a high quality diet (Krause et al. Reference Krause, Honarmand, Wetzel and Naguib2009; Kriengwatana et al. Reference Kriengwatana, Farrell, Aitken, Garcia and MacDougall-Shackleton2015), although spatial memory was impaired (Kriengwatana et al. Reference Kriengwatana, Farrell, Aitken, Garcia and MacDougall-Shackleton2015). In the Panamanian bishop (Brachyrhaphis episcopi), high-predation environments were associated with increased activity and exploration (Archard & Braithwaite Reference Archard and Braithwaite2011). Exposure to environmental stressors such as food availability, or risk of predation, may therefore have some potential benefits in fish and birds, such as increased motivation and exploration. However, a study of foraging demand on bonnet macaques (Macaca radiata) found – perhaps surprisingly – that infants whose mothers had to periodically forage in woodchip were less likely to explore a novel environment than those for whom food was easily available (Andrews & Rosenblum Reference Andrews and Rosenblum1993). This indicates that there may be taxonomic and context-specific differences in how various types of environmental stress are experienced, and that human-controlled environmental deprivation may have different effects to more naturally arising situations such as competition for food.

Environmental conditions summary

The damaging effects of early life stress are extremely relevant to the context of orangutan rehabilitation, as many infant orangutans are rescued from the illegal wildlife trade and found in deprived conditions. In situations where an orangutan is rescued from a human environment, it is reasonable to assume that some detrimental effects of early life stress, e.g. increased anxiety and reduced exploration, may carry through into adulthood, emerging over time. Environmental enrichment can protect against stress, reduce anxiety, depression, and fear, and promote physical activity and spatial memory. However, moving animals from enriched to unenriched environments may have unintended negative effects. This shift in environmental conditions can occur in several circumstances. For example, despite existing enrichment provision in rehabilitation centres, e.g. hammocks and feeding enrichment (Damerius et al. Reference Damerius, Graber, Willems and van Schaik2017), orangutans who have been living in the wild before arriving at a rehabilitation centre will experience a large shift in conditions, from a complex forest environment to the relative deprivation of captivity. However, even within rehabilitation centres there is limited space, and orangutans may need to be temporarily housed in smaller enclosures than those to which they are accustomed. For example, infant orangutans too old for ‘forest school’, i.e. learning skills under human supervision in an area of open forest during the day, may need to be housed for some years in an enclosure before they are old enough to be released into the wild (R Jaya, personal communication 2022). Similarly, orangutans who spend time on a ‘pre-release island’ to experience semi-independent living may be temporarily returned to their enclosure before being released into the wild. In these situations, the regression back to an enclosed environment may have considerably negative effects, including increasing pessimism, which is discussed in the following section. This could be heightened if enclosed environments have negative associations, for example in orangutans who have experienced life in the illegal wildlife trade or who have other prior experience of enclosed environments. Regression to a smaller environment is also likely to reduce levels of physical activity, which in turn is likely to have a negative impact on resilience.

In humans, optimism and pessimism are considered to influence the capacity for resilience; for example, affecting self-perception, environmental perception, and how information is processed and actioned (Forgeard & Seligman Reference Forgeard and Seligman2012). Hobfoll (Reference Hobfoll2002) argues that optimism, self-esteem, and a sense of agency, overlap and tend to be correlated. Optimists can actively approach challenges in a constructive way, perceiving the possibility to act in order to alter outcomes (Forgeard & Seligman Reference Forgeard and Seligman2012). Therefore, an optimistic outlook may help orangutans maintain a sense of control over their surroundings. However, it could also be argued that temporary realistic pessimism may be more appropriate in the face of danger, to prepare for worst-case scenarios and lower expectations of success to prevent disappointment (Forgeard & Seligman Reference Forgeard and Seligman2012). Therefore, a careful balance should be maintained here; the beneficial effects of enrichment can be used in parallel with some constructive environmental challenges that encourage persistence and problem-solving. This requires an understanding of appropriate difficulty levels, so that the challenge presents some degree of frustration, but ultimately elicits a positive emotional state rather than apathy, boredom, or a negative emotional state (Clark Reference Clark2011). In addition, long-term planning is important to ensure that the positive effects are long-lasting and maintained up until the point orangutans are released. It would also be helpful to establish an understanding of current enrichment practices in rehabilitation centres, to measure the effectiveness of different methods and explore the relationship between environmental enrichment and optimism/pessimism in orangutans.